structural enhancement using polyurea

8/12/2019 structural enhancement using polyurea

1/256

STRUCTURAL ENHANCEMENT USING POLYUREA

by

DAVID JAMES ALLDREDGE

A DISSERTATION

Submitted in partial fulfillment of the requirements

for the degree of Doctor of Philosophy

inThe Department of Mechanical and Aerospace Engineering

to

The School of Graduate Studies

of

The University of Alabama in Huntsville

HUNTSVILLE, ALABAMA

2014


2/256

ii

In presenting this dissertation in partial fulfillment of the requirements for a doctoraldegree from The University of Alabama in Huntsville, I agree that the Library of thisUniversity shall make it freely available for inspection. I further agree that permission forextensive copying for scholarly purposes may be granted by my advisor or, in his/herabsence, by the Chair of the Department or the Dean of the School of Graduate Studies. It

is also understood that due recognition shall be given to me and to The University ofAlabama in Huntsville in any scholarly use which may be made of any material in thisdissertation.

____________________________ ___________

(student signature) (date)


3/256

iii

DISSERTATION APPROVAL FORM

Submitted by David Alldredge in partial fulfillment of the requirements for the degree of

Doctor of Philosophy in Mechanical Engineering and accepted on behalf of the Facultyof the School of Graduate Studies by the dissertation committee.

We, the undersigned members of the Graduate Faculty of The University of Alabama inHuntsville, certify that we have advised and/or supervised the candidate on the workdescribed in this dissertation. We further certify that we have reviewed the dissertationmanuscript and approve it in partial fulfillment of the requirements for the degree ofDoctor of Philosophy in Mechanical Engineering.

_________________________________________ Committee Chair(Date)

_________________________________________ Dissertation Advisor(Date)

_________________________________________

_________________________________________

_________________________________________

_________________________________________ Department Chair

_________________________________________ College Dean

_________________________________________ Graduate Dean


4/256

iv

ABSTRACTThe School of Graduate Studies

The University of Alabama in Huntsville

Degree Doctor of Philosophy College/Dept. Engineering/Mechanical andAerospace EngineeringName of Candidate David James AlldredgeTitle Structural Enhancement Using Polyurea

This dissertation explores the potential for using field applied polyurea coatings to

provide structural enhancement to both traditional and non-traditional building materials.

The materials considered include honeycomb composite panels, plywood, lumber, and

cementitious composite panels. The basic approach followed during the project includes

experimental testing, finite element modeling, and a parametric study using the finite

element model(s).

Experimental tests are divided into two distinct series: 1) testing of panel type

materials in four point bending and 2) tension testing of joints in structures fabricated

from lumber. The results obtained from both series of tests show that polyurea can

significantly increase the capacity of the uncoated materials. The research showcases a

unique approach to the strengthening of both traditional and non-traditional building

materials and introduces a potentially game changing technology to the building trade.

Finite element analysis is performed to understand the mechanism in which a

polyurea coating strengthens the materials and to study the impacts of variations to the

relative material properties of the substrate and the coating. The models should help bring

this promising new technology to practice by helping researchers, architects, and builders

select and apply the proper polyurea coating for structural enhancement.


5/256

v

Abstract Approval: Committee Chair

Dissertation Advisor

Department Chair

Graduate Dean


6/256


7/256

vii

first class at UAH. It was an honor to learn from you. Dr Toutanji, you graciously

allowed me to perform the majority of the testing in your labs and I am extremely

grateful for that. It was a privilege to work with you during this work and the additional

work performed under the grant. Dr Lin and Dr Wessling, thank you for your

suggestions and support. Your help in directing the finite element modeling, although

difficult, made the dissertation stronger.

I would like to thank the other members of the research team for their

contributions. Dr Tom Lavin, thank you for your ideas, your time, and guidance. Dr

Madhan (Han) Balasubramanyam thank you for your help in spraying and testing of the

rafter specimens. Hyungjoo Choi, it was an honor to work alongside you. Thank you for

all the time spent in the lab with me. You are such a hard worker and if I worked half as

hard as you, I would have been finished a long time ago. Good luck in completing your

PhD and whatever lies beyond.

I would also like to thank several fellow students for their contribution in

specimen preparation and testing. Your help was invaluable. Jorge Cacciatore, Matthew

Pinkston, Rajesh (Raj) Vuddandam, and Ueno Shigeyuki (Shige) thank you for your help.

Mr. John Becker, president of Creative Material Technologies, thank you for

helping to select the polyureas used in this study and answering all of my numerous

questions. Thank you also for spending the time to teach me the spraying technique.

I want to thank all of my friends. Your sincere interest in my dissertation

motivated me push forward and not to quit. I am blessed to be able to call you friends. I

would also like to thank my coworkers for their support and encouragement. Thanks for

listening to me ramble on about this work.


8/256

viii

To my employer, Dynamic Concepts Inc., thank you for the financial support,

encouragement, and flexibility to go to the lab when I needed to. I would also like to

thank the U.S. Dept. of Commerce for supporting this research under NOAA SBIR Phase

I and Phase II contract No. WC133R-09-CN-0108. Any opinions, findings, conclusions,

or recommendations expressed in this publication are those of the authors and do not

necessarily reflect the views of the Dept. of Commerce.


9/256

ix

TABLE OF CONTENTS

PAGE

LIST OF FIGURES ......................................................................................................... xiv

LIST OF TABLES ............................................................................................................ xx

CHAPTER 1 INTRODUCTION ........................................................................................ 1

1.1 Purpose of the Study ............................................................................ 1

1.2 Motivation for Research ....................................................................... 2

1.3 Objective and Research Plan ................................................................ 3

1.4 Outline of Dissertation ......................................................................... 4

CHAPTER 2 CONSTRUCTION PRACTICES ................................................................. 6

2.1 Current Building Construction ............................................................. 6

CHAPTER 3 POLYUREA ............................................................................................... 13

3.1 Polyurea ............................................................................................. 13

3.2 Polyurea Material Testing .................................................................. 16


10/256

x

3.3 Polyurea Coating Method .................................................................. 24

CHAPTER 4 STRUCTURAL ENHANCEMENT OF BUILDING MATERIALS:PHASE I FLEXURAL TESTS ............................................................................. 26

4.1 Specimen Preparation ........................................................................ 26

4.1.1 Honeycomb Plates ............................................................................. 27

4.1.2 Cementitious Plates ........................................................................... 32

4.2 Testing Methodology ......................................................................... 35

4.3 Results and Discussion ....................................................................... 37

CHAPTER 5 STRUCTURAL ENHANCEMENT OF BUILDING MATERIALS:PHASE II FLEXURAL TESTING METHODOLOGY ...................................... 45

5.1 Specimen Preparation ........................................................................ 45


5.1.2 Plywood Plates .................................................................................. 47


5.2 Testing Methodology ......................................................................... 64

CHAPTER 6 STRUCTURAL ENHANCEMENT OF BUILDING MATERIALS:PHASE II FLEXURAL TESTING RESULTS AND DISCUSSION .................. 66


11/256

xi

6.1 Results and Discussion ....................................................................... 66

6.1.1 Lauan Honeycomb Plates .................................................................. 66

6.1.2 Fiberglass Honeycomb Plates ........................................................... 70

6.1.3 Plywood ............................................................................................. 74


6.2 Conclusion ......................................................................................... 86

CHAPTER 7 STRUCTURAL ENHANCEMENT OF BUILDING MATERIALS:FINITE ELEMENT MODEL DEVELOPMENT ................................................ 88

7.1 Finite Element Model Development .................................................. 88


7.1.2 Plywood ............................................................................................. 93


7.2 Finite Element Model Tuning ............................................................ 99

7.2.1 Lauan Honeycomb Plates ................................................................ 101

7.2.2 Fiberglass Honeycomb Plates ......................................................... 107

7.2.3 Plywood ........................................................................................... 117


12/256

xii

7.2.4 Cementitious Plates ......................................................................... 123

7.3 Finite Element Model Parametric Study .......................................... 136

7.3.1 Lauan Honeycomb Plates ................................................................ 137

7.3.2 Fiberglass Honeycomb Plates ......................................................... 140

7.3.3 Plywood ........................................................................................... 143

7.3.4 Cementitious Plates ......................................................................... 146

CHAPTER 8 STRUCTURAL ENHANCEMENT OF WOOD JOINTS: PHASE ISTRUCTURAL TESTS...................................................................................... 149

8.1 Testing Methodology ....................................................................... 149

8.2 Results and Discussion ..................................................................... 155

CHAPTER 9 STRUCTURAL ENHANCEMENT OF WOOD JOINTS: PHASE IISTRUCTURAL TESTING ................................................................................. 170

9.1 Testing Methodology ....................................................................... 170

9.2 Results .............................................................................................. 175

9.3 Discussion of Results ....................................................................... 176

CHAPTER 10 STRUCTURAL ENHANCEMENT OF WOOD JOINTS: FINITEELEMENT MODEL DEVELOPMENT ............................................................ 183


13/256

xiii

10.1 Finite Element Model Development ................................................ 183

10.2 Finite Element Model Tuning .......................................................... 188

10.3 Finite Element Model Parametric Study .......................................... 201

CHAPTER 11 CONCLUSION AND FUTURE RESEARCH ...................................... 213

11.1 Conclusions ...................................................................................... 213

11.2 Future Work ..................................................................................... 219

APPENDIX A FIXTURE DRAWINGS FOR T SPECIMEN TESTING ...................... 222

APPENDIX B POLYUREA COATING METHODOLOGY ........................................ 224

REFERENCES ............................................................................................................... 229


14/256

xiv

LIST OF FIGURES

PAGE

Figure 3.1 Polyurea Test Setup (a) and Detail of Lower Grip (b) .................................... 18

Figure 3.2 Test Results for 8817 Polyurea (a) and Detail View of Low StressRegion (b) ............................................................................................................. 20

Figure 3.3 Test Results with Average Max Strain and Average Strain (a) andFinal Averaged Results (b) ................................................................................... 20

Figure 3.4 Final Averaged Resuls for 9041 Brush on Polyrea ......................................... 21

Figure 3.5 Final Averaged Results for the 8817 and 9041 Polyureas .............................. 22

Figure 3.6 Voyager Spray System with Polyurea Components and Static Mixer ............ 25

Figure 4.1 Honeycomb with Cloth Face Sheet (a) and Lauan Composite Panel (b) ........ 27

Figure 4.2 Honeycomb Core Placed in Mold (a) and Mix Being Hand Placed (b) .......... 33

Figure 4.3 Example of Completed Panel Before Final Cutting ........................................ 34

Figure 4.4 Schematic of Graphite Mesh (a) and Detail View with Cell Dimensions(b) .......................................................................................................................... 35

Figure 4.5 Four Point Bending Setup with Test Frame (Blue) ......................................... 36

Figure 4.6 Moment Versus Crosshead Displacement for (a) No FiberCementitious, (b) Fiber Cementitious, (c) Fiberglass Honeycomb, and (d)Lauan Honeycomb ................................................................................................ 38


15/256

xv

Figure 4.7 Moment Versus Center Displacement for (a) No Fiber Cementitious,(b) Fiber Cementitious, (c) Fiberglass Honeycomb, and (d) LauanHoneycomb ........................................................................................................... 39

Figure 4.8 Photos and Examples of Nomenclature Used to Describe PolyureaDeposition: Heavily Coated on Front Side (a); Significant Overspray onTop Edge, Regions of Polyurea on Bottom Edge, Front Side (b); Dots ofPolyurea on Both Edges, Front Side (c)................................................................ 43

Figure 5.1 Example of Mold with Placed Plate ................................................................ 55

Figure 5.2 First Attempt at Placing Plates ........................................................................ 56

Figure 5.3 Concrete Specimen Instrumented with Strain Gages ...................................... 58

Figure 5.4 Test Results for Concrete Specimens .............................................................. 59

Figure 5.5 Stress-Strain Tension Plot for Concrete Specimen 1 ...................................... 60

Figure 5.6 Stress-Strain Compression Plot for a Similar PVB Mix[47] ........................... 63

Figure 5.7 MTS Loading Frame with Fixture and Loaded Specimen .............................. 65

Figure 6.1 Lauan Honeycomb Panel Moment-Displacement Test Results ...................... 67

Figure 6.2 Shear Failure of Lauan Panel .......................................................................... 69

Figure 6.3 Fiberglass Honeycomb Panel Moment-Displacement Test Results ................ 71

Figure 6.4 First (a) and Second (b) Failures of the Coated Top 1 Specimen ................... 72

Figure 6.5 Plywood Moment-Displacement Test Results ................................................ 75

Figure 6.6 Cementitious Panel Moment-Displacement Test Results ............................... 81


16/256

xvi

Figure 7.1 Four Point Bending Finite Element Model ..................................................... 89

Figure 7.2 Low Tension Material Stress-Strain Diagram[53] .......................................... 98

Figure 7.3 Comparison of Lauan Tests and Initial Finite Element Model Results......... 102

Figure 7.4 Uncoated Lauan Test Results with FEM Comparison at PredictedFailure ................................................................................................................. 103

Figure 7.5 FEM Results for Lauan Uncoated and Coated Configurations ..................... 105

Figure 7.6 Comparison of Fiberglass Tests and Initial Finite Element ModelResults ................................................................................................................. 108

Figure 7.7 Fiberglass Results with Initial FEM and First Tuned FEM .......................... 109

Figure 7.8 Fiberglass Test Results Compared to Tuned Finite Element Models ........... 110

Figure 7.9 Finite Element Contour Plot of Outer Layer Tensile Stresses for theFiberglass Plate ................................................................................................... 111

Figure 7.10 Uncoated Fiberglass Panels Test Results .................................................... 112

Figure 7.11 Uncoated Fiberglass Test Results with FEM Comparison at PredictedFailure ................................................................................................................. 115

Figure 7.12 FEM Results for Fiberglass Uncoated and Coated Configurations............. 116

Figure 7.13 Comparison of Plywood Tests and Initial Finite Element ModelResults ................................................................................................................. 118

Figure 7.14 Uncoated Plywood Results with FEM Results at Predicted FailureBefore Updating .................................................................................................. 120

Figure 7.15 FEM Results for Plywood Uncoated and Coated Configurations ............... 121


17/256


18/256

xviii

Figure 8.11 When the Orientation of the Grain in the Lower Member of the TopPlate was Horizontal (b), the Joint was Stronger. ............................................... 161

Figure 8.12 Load Versus Deflection Plots for the Uncoated Reinforced Standardand Coated Reinforced Configurations. .............................................................. 162

Figure 8.13 The Coated Reinforced Configurations Failed Relatively Slowly (a)or Quickly (b) when Wood Fibers Fractured in the Top Plate. .......................... 162

Figure 8.14 Load Versus Deflection Plots for 2x4 End Nail Failure in SixDifferent Configurations ..................................................................................... 164

Figure 8.15 A Partial Configuration is Retested for End Nail Failure............................ 166

Figure 8.16 A Partial Coated Configuration Was Retested (a) to Evaluate Toe-Nail Rafter Failure and a Crack Developed in the Rafter (b) ............................. 167

Figure 8.17 Load Versus Deflection Plot for Toe-nail Rafter Failure of theUncoated Unreinforced Standard and Unreinforced Configurations Coatedwith Black and White Polyurea .......................................................................... 168

Figure 9.1 T Specimen Layout and Dimensions ............................................................. 172

Figure 9.2 Test Setup for T Specimens Showing the Test Fixture ................................. 173

Figure 9.3 Example of Higher Loaded Glued Specimen (a) and Lower LoadedSpecimen (b) ....................................................................................................... 179

Figure 9.4 Coated with Vertical Grain Specimen 3 During Testing ............................... 181

Figure 10.1 Finite Element Model of T Specimen (a) and Detail View ShowingCoating (b) .......................................................................................................... 184

Figure 10.2 Force-displacement Results for the Uncoated Fine Grain Specimens ........ 187

Figure 10.3 Force-displacement Results for the Uncoated Fine Grain SpecimensCompared to FEM Results .................................................................................. 188


19/256

xix

Figure 10.4 Detail View of Glued T Specimen FEM ..................................................... 189

Figure 10.5 Horizontal Grain Glued Specimen Test and FEM Results .......................... 190

Figure 10.6 Details of T Specimen Assembly into Fixture ............................................ 192

Figure 10.7 Hoffman and Maximum Stress Criteria Envelope Curves .......................... 194

Figure 10.8 Stress Distribution at Interface for Glued FEM .......................................... 195

Figure 10.9 Detail of Nominal Thickness Coated FEM ................................................. 196

Figure 10.10 Coated FEM Showing Failed Wood Elements (Red)................................ 200

Figure 10.11 Capacity Versus Throat Length for the Coated T Specimens ................... 204

Figure 10.12 Coating Technique Used During Research (Left) and TargetedSpraying Indicated by FEM Results (Right) ....................................................... 205

Figure 10.13 Stress-Strain Plots for Polyurea Variations (a) and Zoomed Plot (b) ....... 207

Figure A.1 Angle Bracket for T Specimen Testing ........................................................ 222

Figure A.2 Grip Plate ...................................................................................................... 223

Figure B.1 Voyager Spray System Components ............................................................ 225


20/256

xx

LIST OF TABLES

PAGE

Table 3.1 Examples of Commercially Available Polyurea ............................................... 15

Table 3.2 Polyurea Properties ........................................................................................... 16

Table 3.3 Hanging Weight Test Results for the White and Black Polyurea..................... 17

Table 4.1 Honeycomb Face Sheet Material Properties ..................................................... 28

Table 4.2 Material Properties of Polypropylene ............................................................... 29

Table 4.3 Summary of Honeycomb Core Properties ........................................................ 31

Table 4.4 Bending Specimen Test Results ....................................................................... 40

Table 4.5 Notes Regarding Polyurea Deposition and Failure Mode. ............................... 41

Table 5.1 Number of Specimens per Configuration ......................................................... 46

Table 5.2 Minimum Structural Plywood Requirements ................................................... 49

Table 5.3 Mechanical Properties of Solid Sawn Pine Wood ............................................ 51

Table 5.4 Structural Properties of Plywood ...................................................................... 52

Table 5.5 High-performance Cementitious Mix Design [41] ........................................... 54

Table 5.6 Raw Fiber Material Properties .......................................................................... 57

Table 5.7 Material Properties for PVB Concrete .............................................................. 59


21/256

xxi

Table 5.8 Summary of Test Data for PVB Concrete ........................................................ 64

Table 6.1 Summary of Lauan Panel Results ..................................................................... 68

Table 6.2 Summary of Fiberglass Panel Results .............................................................. 73

Table 6.3 Summary of Plywood Panel Results ................................................................. 75

Table 6.4 Comparison of Moment of Inertia Based on Outer Layers .............................. 79

Table 6.5 Summary of Cementitious Panel Results.......................................................... 83

Table 6.6 Summary of Normalized Cementitious Panel Results ...................................... 84

Table 7.1 Error in Stress Calculation for a Given Number of Layers .............................. 91

Table 7.2 Layup Properties for Lauan Reinforced Honeycomb Plates ............................ 93

Table 7.3 Layup Properties for Fiberglass Reinforced Honeycomb Platess .................... 93

Table 7.4 Layup Properties for Pine Plywood Plates ....................................................... 94

Table 7.5 Summary of Material Cementitious Plate FEM Properties .............................. 97

Table 7.6 Layup Properties for Cementitious Plates ........................................................ 97

Table 7.7 Example Layup of Top and Bottom Coated Lauan ........................................ 100

Table 7.8 FEM Predicited Shear Stress at Honeycomb-Lauan Interface ....................... 104

Table 7.9 FEM Predicted Moments at Failure for Lauan Panels .................................... 106

Table 7.10 FEM Predicited Shear Stress at Honeycomb-Fiberglass Interface ............... 113


22/256

xxii

Table 7.11 Plywood FEM Moments at Predicted Failure............................................... 119

Table 7.12 Layup Definition for Modified Cementitious Plate ...................................... 125

Table 7.13 Description of FEM Modifications for Cementitious Plates Tuning ............ 126

Table 7.14 Additional FEM Modifications for Cementitious Plates Based on FEMMod 4 .................................................................................................................. 129

Table 7.15 Failure Indicators for Tuned Cementitious FEM .......................................... 134

Table 7.16 Coated FEM Results for Cementitious Plates............................................... 135

Table 7.17 Parameter Variations for Lauan Honeycomb Panels .................................... 137

Table 7.18 Summary of Results for Lauan Parametric Study ........................................ 139

Table 7.19 Parameter Variations for Fiberglass Honeycomb Panels ............................. 141

Table 7.20 Summary of Results for Fiberglass Parametric Study .................................. 142

Table 7.21 Parameter Variations for Plywood Plates ..................................................... 143

Table 7.22 Summary of Results for Plywood Parametric Study .................................... 145

Table 7.23 Parameter Variations for Cementitious Plates .............................................. 146

Table 7.24 Summary of Results for the Cementitious Parametric Study ....................... 147

Table 8.1 Configurations Tested During Phase I Structural Tests.................................. 153

Table 8.2 Tabulated Results for 2x4 End Nail Failure in Six DifferentConfigurations..................................................................................................... 165


23/256

xxiii

Table 8.3 Tabulated Results for Toe-nail Rafter Failure in Three DifferentUnreinforced Configurations .............................................................................. 169

Table 9.1 T Specimen Configurations ............................................................................ 174

Table 9.2 Summary of Test Results for Phase II Structural Testing .............................. 175

Table 10.1 Summary of FEM Results for Coated T Specimens ..................................... 198

Table 10.2 T Specimen FEM Results of Additional Thickness Variations .................... 202

Table 10.3 Results of Increased Throat Length Variation Study .................................... 203

Table 10.4 Variations of the Mechanical Properties of 8817 Polyurea .......................... 206

Table 10.5 Results of Polyurea Properties Variation Study ........................................... 208

Table 10.6 Summary of Nail Variations for Parametric Study ....................................... 210

Table 10.7 Results of Parametric Study for Nail Properties ........................................... 212


24/256


25/256

2

1.2 Motivation for Research

In a 2009 report about the devastating Hurricane Ike, the Insurance Institute for

Business and Home Safety (IBHS) states that there are more than $9 trillion worth of

insured properties along the Gulf and Atlantic coasts and as much as 50% of the US

population resides within 80 km of the coast [2]. It is obvious that with such a high

population and property value exposed to coastal environmental risks that there would be

the need for ongoing research into materials and designs that increase the likelihood a

structure could withstand natural hazards. Overall, the results contained herein indicate

that the commercial applications of using polyurea to strengthen structures in hurricane

prone areas could be enormous.

The IBHS was formed to help reduce the monetary and human costs by providing

scientific research to identify and promote effective actions that strengthen homes,

businesses, and communities against natural disasters and other causes of loss. [3] The

research has resulted in recommendations for home and business owners. These

recommendations range from securing picture frames for earthquake protection to ways

to strengthen traditional construction techniques for protection against hurricanes.

However, no mention is made by this agency, and no evidence could be found elsewhere,

regarding the use of field applied polyurea coatings to provide an improved and

continuous foundation to roof load pathway. Therefore, the work reported herein

represents a unique, and arguably significant, contribution to the field.


26/256

3

1.3 Objective and Research Plan

The objective of this research is to determine if using a polyurea coating on

traditional and non-traditional building materials can increase the load capacity of the

materials. The materials to be considered include honeycomb composite panels,

plywood, lumber, and cementitious composite panels. Plywood, honeycomb panels, and

cementitious panels can be used as both sheathing and roof decking.

One important goal of this research is to determine if the application of polyurea

can increase the strength of the base material. With respect to sheathing and decking, this

has implications for resistance to both wind produced lateral loads as well as impact

resistance to flying debris. The application of polyurea onto lumber, which is the primary

choice for framing in residential structures, has implications for strengthening the

continuous load path from roof to ground.

The basic approach followed during the project includes experimental testing,

finite element modeling, and a parametric study using the finite element model(s). The

experimental tests are divided into two distinct series: 1) flexural testing of panel type

materials in four point bending and 2) structural testing of joints fabricated from lumber.

Specifically, results from Phase I feasibility tests are used to direct Phase II testing.

Finite element models are created and tuned to match the test results. The tuned finite

element models are then used to conduct a parametric study to test the sensitivity of the

configurations to variables like polyurea thickness, strength, etc.


27/256


28/256

5

where the sensitivity to parameters, such as coating thickness and material properties, are

determined.

Chapter 8begins the portion of the document concerned with the structural

enhancement of lumber using polyurea. First in Chapter 8, the Phase I structural test

program is documented. Phase I was conducted to determine if a polyurea coating could

strengthen the rafter-to-top plate connection. Results and observations are presented and

these results and observations are used to direct the Phase II structural testing program.

Phase II structural testing of polyurea coated lumber is described in Chapter 9.

The chapter begins with a description of the specimen preparation and testing

methodology. Next the results of the testing program are presented. The chapter

concludes with a discussion of the test result.

Chapter 10details the development of the finite element models of the Phase II

testing configurations. First, the baseline model is described along with the necessary

tuning needed to best match the uncoated test results. Next the development and results

of the coated models are presented. Lastly, a parametric study is conducted and detailed.

Chapter 11brings the dissertation to a conclusion and includes suggestions for

future research.


29/256

6

CHAPTER 2

CONSTRUCTION PRACTICES

This chapter focuses on current construction practices in terms of a review of the

most prominent building code used in the United States. This is followed by field

observations following major weather events that highlight the fact the homes built per

code perform well in these events. Unfortunately, not all homes are built to withstand

high wind events either because the code is not required based on location or because the

code was not followed.

2.1 Current Building Construction

Building codes like the International Building Code (IBC) and the International

Residential Code (IRC) have been adopted by many states and local governments in an

attempt to standardize residential and commercial construction. Codes provided builders

with a set of minimum requirements that must be met in order to pass inspection.

Although the IBC and IRC have been widely adopted, in many states neither code has

been adopted at the statewide level for all buildings. Both the IRC and IBC are

considered to be prescriptive codes that are based upon engineering analysis and design

manuals. Here, a prescriptive code is one in which the minimum building requirements

are set forth and for which no engineering calculations are required.


30/256

7

For the majority of single family residential construction, the IRC is the

applicable design document. In order to prescribe building requirements for wind

resistance, the IRC provides a map of the U.S with the applicable design wind speeds.

As expected, the coastal areas of the US require resistance to higher wind speeds. For

areas of the US where the design wind speeds are greater than 160 km/h, or in specially

designated high wind areas, the IRC requires the structure to be designed to more

stringent codes like the AF&PA Wood Frame Construction Manual (WFCM) and the

ICC Standard for Residential Construction in High-Wind Regions (ICC 600). These

manuals and codes are not prescriptive and require engineering design and analysis.

Both the IBC and IRC are published by the International Code Council (ICC),

which was established in 1994 by the developers of the three main regional building

codes. The formation of the ICC was precipitated by the desire to have one national

building code without regional limitations [4]. The first IBC code published by the ICC

was in 2000, which was about the same time that the state of Florida adopted its first

single statewide building code.

Of the 96 major hurricanes to hit the mainland U.S coast since 1851, 37 have

impacted the state of Florida, which is nearly double the number of strikes as the next

state [5]. In the 1970s,the first law mandating the adoption and enforcement of one of

four state recognized minimum building codes was enacted [6]. After major hurricanes

in the early 1990s, the state of Florida reviewed the state building code and found that

adoption and enforcement was inconsistent throughout the state and those local codes

thought to be the strongest proved inadequate when tested by major hurricane events. [6]

This led to legislation in 1998 that established the Florida Building Commission and


31/256

8

instructed the Commission to develop a statewide code. The Commission finalized and

submitted a draft to the Florida Legislature in 2000 and the code went into effect in

January of 2001 [6]. The Florida building code is mainly adapted from both the IBC and

IRC set of codes.

As stated previously, the IRC provided a set of maps showing design wind

speeds. Per the provided maps, zones that require high wind design are located within

approximately 240 km of the coast for the portion of the U.S from Texas to Maine. This

means that most new construction within these areas is required to have a continuous load

path from roof-to-ground. The converse is also true, which means that in areas further

away from the cost no additional connectors are required at the rafter-to-top plate

connections other than the toe nailed connection. This makes sense for protection against

hurricanes, but many of the areas not required to strengthen the roof connections lie in

tornado prone areas.

North and Central Alabama are areas that are outside the higher wind zones in the

IRC but are prone to tornados. In fact, in late April of 2011, a historic outbreak of severe

weather occurred along the Southeastern U.S. that produced a total of 62 tornados in

Alabama alone [7]. Widespread destruction and loss of life were reported throughout the

state and region. Tornados have the ability to produce winds in excess of 400 km/h, be

over 1.6 km wide and can remain on the ground for up to 80 km [8]. Damage to homes

and businesses include roof damage, roof removal, wall collapse, and complete

destruction. Although the highest winds are located at the near the center of tornado,

damage and destruction does occur along the outer portions of the storm where winds are


32/256

9

lower. This was evident in a post-tornado report on the storms that moved through

Tuscaloosa, AL as a part of the outbreak in 2011.

Researchers from several universities and interested companies, funded by the

National Science Foundation (NSF), surveyed the aftermath of the tornados in

Tuscaloosa in a focused and methodical manner in order to document the damage and

failure modes in primarily wood-frame construction.[9] They spent three days

surveying the damage along the tornado path, both near the assumed center of the storm

as well as the areas away from the center.

The conclusion reached by the team was that light-frame wood buildings do not,

and will not, have the ability to resist EF4 or EF5 tornadoes. [9] They add that a

majority of the damage to residential construction occurs at wind speeds below the EF

rating and that most of the buildings along the path of a strong tornado, even along the

outer edge, are not repairable based on the current construction techniques. This, they

say, provides an opportunity and incentive for tornado resistant construction practices,

which do not exist currently.

The researchers documented several case studies along the path of the tornado and

the damage and failure methods associated with each. Some examples of the damage

they noted included roof failures where the rafters were toe nailed to the top plate, houses

that shifted off the foundation due to improperly installed or missing anchor bolts, and

damage to lateral walls and gable roofs from lateral wind pressures. The researchers

propose a design philosophy to reduce monetary losses and increase safety.

From a monetary standpoint, the design solution the team proposes is meant to

shrink the total damage footprint and the severity of the damage toward the center of the


33/256

10

storm. This is accomplished by design improvements on a component level (i.e., roof

decking) as well as a system level (i.e., entire roof). As related to the current research,

the authors recommend a continuous load path from roof-to-ground and increased shear

resistance of lateral walls by means of nail spacing and anchoring. In addition to design

solutions, the authors also realize that much more research is needed to understand the

unique loading scenario that tornadoes produce when compared to more straight line

wind events like hurricanes.

As previously stated, building codes such as the IRC and the Florida Building

Code have been addressing hurricane resistant construction since the early 2000s. Post

hurricane insurance assessments made after Hurricane Charley in 2004 indicated that

homes built to the more stringent codes had a reduced number of claims with less severe

damage. Additionally, when damage did occur, homeowners were able to return more

quickly to their home if the structure was built to updated building codes [10].

Unfortunately, newer building codes do nothing to protect previously built

structures and newer homes required to be built per the more stringent codes are not

guaranteed to actually follow the code. Enforcement of the building codes are the role of

building inspectors. This is an arduous task considering the likelihood that the local

inspection department is under staffed and under resourced coupled with the amount of

requirements in the building codes that must be inspected. With regards to the uplift

capacity of the structure, consider the sheer number of nails that must be installed

properly to obtain the desired capacity. Proper installation of the nails, which in most

cases cannot possibly be fully inspected, includes the number of required nails, the


34/256

11

spacing, the size, and in some cases the type of shank on the nail. Post Hurricane Katrina

assessments of residential structures highlight the importance of adherence to the codes.

Following Katrina, researchers spent 3 days along the Gulf Coast gathering data

on residential wood frame structures with the ultimate goal of providing the pertinent

data necessary to improve the performance of wood structure during high wind events

[11]. Their findings indicate that the structures built to the updated codes performed well

but that in many cases the current codes requirements were not met. For example, they

noted in numerous cases that the nail spacing on roof decking failed to meet the

minimum requirements and resulted in loss of sheathing. Additionally, they found

improper number of nails in a hurricane strap and improper anchoring of a top plate to the

wall. These shortcomings led to the loss of a roof in a condominium community.

Missing nails is one of the most common mistakes that Jim Mattison of Simpson Strong-

Tie sees in the field [12].

Mattison states that clips without the proper number of nails cannot handle the

loads they were designed for. This can lead to rotation of the clip, which can damage the

adjacent wood, or it can lead to clip failure. Besides the incorrect number of nails, he

states that he has observed instances where the installer used the wrong size and type of

nails in an effort to save money. Smaller diameter nails result in a lower shear capacity

and shorter nails have a lower resistance to withdrawal. Lastly, Mattison points out that

the use of pneumatic nail guns can be problematic. He has found cases where nails made

their own hole instead of the factory-punched hole as well as cases of overdriven nails

that result in excessive dimpling. In both cases, the load capacity of the hurricane clip is

diminished.


35/256

12

The goal of this research is to determine if polyurea can strengthen typical

building construction to withstand natural disasters including, but not limited to,

hurricanes and tornados. Wind and windblown debris are one of the main hazards

experienced during both hurricanes and tornados. Wind pressures can produce high loads

on lateral walls as well as high uplift forces. The high lateral loads can cause failure of

the impinged wall as well as a racking failure of the supporting walls. Uplift forces, if

not resisted from roof to ground, can cause failures like roof removal, wall collapse, and

shifting of the structure off of the foundation. Applications for the polyurea

reinforcement may include new construction and existing construction to replace

typically used reinforcement, to enhance already installed reinforcement, or to strengthen

structures without any required reinforcement.


36/256

13

CHAPTER 3

POLYUREA

This chapter describes polyurea and its use in this research. An introduction to

polyurea including a brief history is presented first. This includes some uses of polyurea

as well as material properties for several different commercially available types. Next,

material testing of the polyurea formulation used in the current research is described.

This testing was performed to determine material properties that were needed for finite

element modeling. Last, a step by step guide to the spraying method is described.

3.1 Polyurea

Polyurea is a two part polymer with a rapid gel time that results in a 100% solid

coating containing zero volatile organic compounds (VOCs). As a coating, polyurea has

been used to protect against corrosion, moisture, abrasion, and chemicals in a variety of

different applications. Some examples of typical applications using polyurea include

truck bed liners, pond liners, concrete floor coating, water tank liners, commercial

roofing, and pipeline corrosion protection. Polyurea has also shown the ability to protect

occupants in buildings and vehicles from debris and spall from blast waves.

The Polyurea Development Association (PDA) defines pure polyurea as the

reaction of a polyisocyanate component and an amine-terminated resin blend [13]. The


37/256

14

isocyanate component may be either aliphatic or aromatic. The resin must be made up

of amine-terminated polymer resins, and/or amine-terminated chain extenders.[13] By

contrast, polyurethane is made by the reaction of an isocyanate component and a

hydroxyl terminated resin. In addition, polyurethane requires the use of a catalyst in order

to facilitate the chemical reaction whereas polyurea does not [14].

The first reference to polyurea came in 1948 when researchers discovered that

these compounds had far superior thermal properties and extremely high melting points

compared to other polymer systems such as polyesters, linear polyethylene,

polyurethanes, and polyamides [15]. The high thermal stability eventually led to the use

of polyurea in the Reaction Injection Molding (RIM) process used to manufacture

automobile body panels and fascia [16]. The high thermal stability of polyurea when

compared to other polymer systems allowed for the use of high temperature painting

techniques that would damage parts made from polyurethane and other polymer systems

[14].

The biggest challenge in moving polyurea from the RIM process to a coating

system was advancing the spray equipment to be able to cope with the rapid gel time of 1

2 sec without compromising the unique characteristics of the material. Although work

was done in the 1970s with modified polyamines and high levels of plasticizers and

solvents in order to achieve a spray system for coating work [17], poor field performance

was noted and this technology never gained acceptance. By the late 1980s however,

spray equipment had advanced such that polyurea could be used in the coating industry

[18].


38/256

15

Those early polyurea formulations resulted in gel time of approximately 2 sec,

tensile strength near 14 MPa, and elongation around 200% [18]. Generally the polyurea

formulation was a derivative of the polyurea used in the RIM process and was sprayed at

a high temperature and pressure. Since that time, polyurea manufacturers have been able

to produce polyurea with a wide variety of properties and application methods. Table 3.1

lists some examples of commercially available polyurea formulations and their respective

properties. Although the table only represents a small portion of the available polyurea,

it does highlight the variance in mechanical properties and application techniques. The

polyurea selected for this research is detailed in Section3.2.

Table 3.1 Examples of Commercially Available Polyurea

PropertyDragonshield-

HT ERC [19]Watershield

III [20] FSS 45DC [21]

X-Shield

Patch Coat

[22]

Tensile Strength [MPa] 29.09 >18.07 13.44 - 16.20 11.72

Elongation [%] 619 930 450-520 45

Hardness [Shore D] 44-52 84 45 50

100% Modulus [MPa] 8.83 3.68 6.62 -

Gel Time 6 sec 10 sec 20-30 sec 5 min

Spray Temp [C] 79 71-76 77 Brushable

Spray Pressure [MPa] 20.7 13.8 13.8 -

In addition to the properties listed inTable 3.1,polyurea has been shown to have

excellent adhesion to a wide variety of substrates, good flexibility at low temperatures,

and high toughness [23]. Researchers have also determined that certain polyurea can be

highly strain rate sensitive and show significant strain hardening [24]. These properties


39/256


40/256

17

For purposes of comparing the two sprayable formulations, hanging weight tests

were performed on both formations during the Phase I structural tests. Strain gages were

utilized to determine the elastic modulus and the Poissons ratio for the 8817 and the

1137 polyureas. Strain gages were bonded to both the front and back surfaces in the

longitudinal and lateral directions. The specimens, produced by Creative Materials, were

measured to be approximately 2.5 cm wide with a thickness of 2 mm for the 8817 (white)

and 5 mm for the 1137 (black). The tests were conducted by suspending the specimens

from a grip of a loading frame. A weight hanger was hung from the free end of the

specimen using a hole that was drilled through the specimen. Weights were added to the

hanger 100g at a time up to 900g and the strain was recorded at each load increment.

Using the measured dimensions and the hung weight, the maximum applied stress was

167.5 kPa for the white and 68.7 kPa for the black. The elastic modulus and Poissons

ratio was determined by using a best fit curve through the average strain from both the

longitudinal and lateral strain gages. The results, which are shown below inTable 3.3,

show that the modulus for the 8817 polyurea is nearly 3 times that of the 1137 and that

both formulations have relatively high Poissons ratios.

Table 3.3 Hanging Weight Test Results for the White and Black Polyurea

Polyurea Elastic Modulus (MPa) Poisson's Ratio

1137/Black 179 0.52

8817/White 480 0.4

As stated above, the hanging weight tests were performed for the purposes of

comparing the modulus of the 8817 and 1137 formulations. For the purposes of the


41/256

18

Phase II flexural and structural tests, and more specifically the finite element modeling, a

full characterization of the stress-strain history was needed for the 8817 and the 9041

(brushable). A set of tensile tests were performed to obtain the material properties for

both of these polyureas. For each, a set of four specimens were fabricated and tested to

failure. The specimens were produced by first pouring the polyurea into a rectangular

mold, then allowing them to cure. After curing, 25.4 mm wide specimens were cut from

the hardened sheets. The thickness of the specimens varied between formulations with

the thickness across both formulations ranging from 1.88 mm to 5.33 mm. Digital

calipers were used to measure the width and thickness of each specimen. A total of 3

width measurements and 6 thickness measurements were taken to determine the average

cross sectional area, which was used in the stress calculations. The test setup is shown in

Figure 3.1a.

(a) (b)

Figure 3.1 Polyurea Test Setup (a) and Detail of Lower Grip (b)


42/256

19

The tension tests were conducted using a 44 kN Satec (now known as Instron)

load frame equipped with a load cell and capable of measuring load and displacement.

The tests were all performed using the displacement control setting with a pull rate of

12.7 mm/min. Prior to running the tests, the gauge length for each specimen was

measured as the distance between the grips. This value was used along with the recorded

displacement to determine the strain in the specimen. The measured force time history

was used along with the average cross sectional area to determine the stress in the

specimen.

The results from the 8817 (white) polyurea are shown inFigure 3.2a. A detail

view of the early portion of the test is also shown inFigure 3.2b. Examination of the

detail view reveals a nearly horizontal portion of the stress-strain curve near 500 kPa. At

first, it was speculated that this was the result of the specimen slipping in the grips. After

further inspection, it was found that all specimens had this feature and that it always

occurred at approximately the same force level, which seemed unlikely to be from

slipping for all specimens. It was determined that the actual source of the increase in

displacement was the movement of the lower grip in the mount mechanism. The lower

grip, under gravity, rested on the top of the mount, but was being restrained by the pin

when the vertical load exceeded the weight of the grip (Figure 3.1b). The test results

were therefore modified to remove this motion which appeared as a sudden change in

strain.


43/256

20

(a) (b)

Figure 3.2 Test Results for 8817 Polyurea (a) and Detail View of Low Stress Region (b)

In order to utilize the test results in a finite element model, the results were

averaged to produce a characteristic curve for each formulation. The first step in

producing the average curve was to calculate the average breaking strain. This was

selected as the strain just prior to the rapid decrease of stress. The selected values are

represented as circles and the average of the four values is indicated by the vertical light

blue line inFigure 3.3a.

(a) (b)

Figure 3.3 Test Results with Average Max Strain and Average Strain (a) and

Final Averaged Results (b)

The second step was to average the four curves in a strain region where all four

specimens had not failed. A strain of 0.55 mm/mm was selected and the average curve

0 0.2 0.4 0.6 0.8 1 1.2 1.40

5

10

15

20

25

Strain [mm/mm]

Stress

[MPa]

Specimen 1

Specimen 2

Specimen 3

Specimen 4

0 0.02 0.04 0.06 0.08 0.1 0.12

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

5.5

Strain [mm/mm]

Stress[MPa]

Specimen 1Specimen 2


0 0.2 0.4 0.6 0.8 1 1.2 1.40

5

10

15

20

25

Strain [mm/mm]

Stress[MPa]


Specimen 3

Specimen 4

AverageAverage Max Strain

0 0.2 0.4 0.6 0.8 1 1.2 1.40

5

10

15

20

25

Strain [mm/mm]

Stress[MPa]

Specimen 1

Specimen 2

Specimen 3

Specimen 4

Average


44/256

21

can be seen inFigure 3.3a. The last step was to use the MatlabPolyfitfunction to fit a 2nd

order polynomial to the average strain curve from a strain of 0.3 mm/mm to 0.55 mm/mm

and then extrapolating the polynomial from 0.55 mm/mm to the average breaking strain

(vertical light blue line).

The final results of the method are shown inFigure 3.3b and are labeled

Average. One note is that the peak stress of the final averaged results does not match

the average of the maximum stress from the 4 test specimens. In the case of the 8817, the

peak of the final averaged results is 16.93 MPa versus an averaged max stress of 17.83

MPa.

Similar to the 8817, the tension test results for the 9041 brush on polyurea were

processed and averaged. Both the test results and the final averaged results are shown in

Figure 3.4.

Figure 3.4 Final Averaged Resuls for 9041 Brush on Polyrea

0 0.5 1 1.5 2 2.50

1

2

3

4

5

6

7

8

9

10

11

Strain [mm/mm]

Stress[MPa]

Specimen 1

Specimen 2

Specimen 3

Specimen 4

Average


45/256

22

The average results for the 8817 and 9041 are plotted together inFigure 3.5 for

comparison. It is clear that the 8817 polyurea is much stronger than the 9041 and that the

modulus of the 8817 is also higher. An elastic modulus and yield stress were determined

from the averaged stressstrain plots. The values were determined by fitting a linear

curve to initial portion of the curve. For the 8817 and the 9041, the modulus was

determined to be 106 MPa and 72 MPa respectively. The yield strength was determined

to be 13.8 MPa for the 8817 and 4.1 MPa for the 9041. These values are explicitly

needed in the finite element representation of the polyurea (see Chapters 7 and 10).

Figure 3.5 Final Averaged Results for the 8817 and 9041 Polyureas

Comparing the elastic modulus of the 8817 from the hanging weight test and the

tension test shows a large difference in the modulus. The hanging weight tests result in a

much higher modulus than that determined from the tension tests. The specimens and

results were inspected and it was concluded that the data from the tension tests was the

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.80

2

4

6

8

10

12

14

16

18

20

Strain [mm/mm]

Stress[MPa]

8817 White

9041 Brush On


46/256

23

most appropriate to use. As stated above, the manufacturer produced the hanging weight

specimens by means of spraying a coating over a nonstick surface. This resulted in a

specimen with one smooth surface and one severely dimpled surface. The dimpled

surface made it difficult to determine the appropriate thickness to use for the stress

calculations. Also with this specimen, there were areas of black speckle that appeared

along the centroid. The manufacturer stated that this was deemed burning and was the

result of the heat generated due to the thickness of the specimen. The specimens

manufactured by the author for the tension tests were poured, not sprayed, which resulted

in two smooth surfaces and showed no burningeven thought the specimens were

thicker than the hanging weight specimen. In fact, the black burned areas were not seen

in any of the testing performed during this research whether sprayed or poured. It is

unclear what impact, if any, the burning would have on the mechanical properties

derived from the hanging weight test.

In addition to the above, the test procedure was examined for differences. Since

the elastic modulus was the desired result of the hanging weight test, the test was only

performed to stress level of ~170 kPa which is over 100 times lower than the ultimate

strength of the material. Loading of the specimens was low since the results of the

hanging weight tests were used to simply make observations regarding the differences of

the 8817 and 1137 polyureas. It is unclear that the modulus of the hanging weight

specimen would have remained the same at the higher stress-strain levels of the tension

tests.

For the tension tests, it was expected for the polyurea to fail at strain levels at or

above 100% strain based on discussion with the manufacturer and through searching of


47/256

24

the literature. Based on this, it was decided to use the cross head displacement in order to

determine the strain since the typical strain gage is valid for strains up to ~5%. It is

possible that slipping of the specimens within the grips occurred although inspection of

the specimens after testing did not indicate any slipping. In retrospect, it would have

been useful to utilize strain gages during the tension tests to help validate the strain

determined from the cross head displacement. Full characterization of the material with

validated results will be left to future research.

With only the stress-strain data determined by the cross head displacement, the

open literature was searched for comparable test results in order to validate the results. It

was found that very little test data has been published for quasi-static testing. A majority

of the published test results are for high strain rates which is not applicable to the current

research. Three sources of data were found in the literature and the reported modulus for

low strain rates ranged from 49.5 MPa to 192 MPa, which compares well with the

modulus calculated using the cross head displacement [27][28][29]. Certainly the cited

results are for different formulations of polyurea but the modulus of the 8817 lies within

the range. Based on this and the investigation of the specimens and results, the properties

obtained during the tension tests will be used, but modification of the properties will be

parametrically studied during the finite element development.

3.3 Polyurea Coating Method

Creative Materials develops polyurea that can be sprayed at pressures at or

below 415 kPa by means of one of the Voyager low pressure sprayers (Error! Reference

source not found.). For this research, the low pressure cartridge system was used which


48/256

25

utilizes two 750 ml cartridges in a 1:1 mix ratio along with a static mixer to properly mix

the A and B sides. The coating application method is detailed in Appendix B.

Figure 3.6 Voyager Spray System with Polyurea Components and Static Mixer


49/256

26

CHAPTER 4

STRUCTURAL ENHANCEMENT OF BUILDING MATERIALS:

PHASE I FLEXURAL TESTS

Traditional and non-traditional building materials were tested in order to

determine if polyurea could be used to increase the flexural performance of the materials.

This chapter describes the testing of these materials and results of the tests. The materials

tested included two types of honeycomb composite panels and two reinforced

cementitious panels. Specimens were prepared and tested in four point bending to assess

the bending strength of the materials in both uncoated and polyurea coated

configurations.

4.1 Specimen Preparation

Four flexure specimens from four different materials were fabricated for a total of

16 specimens that each had dimensions of 61 cm X 10.2 cm with varying thickness

between the materials. Two of the materials tested were commercially available

honeycomb panels. The first was Nida-Core H11PP honeycomb core with 18 oz

fiberglass face sheets and the second was Nida-Core H11PP honeycomb panel with

Lauan face sheets. Both materials were purchased in large panels and simply cut to size

using a table saw. The other two panels tested were graphite reinforced cementitious


50/256

27

panels one with and one without free fibers. The graphite reinforced cementitious panels

were fabricated by placing two layers of graphite mesh over 13 mm thick Nida-Core

H11PP honeycomb core that was filled with one of the cementitious matrixes.

4.1.1 Honeycomb Plates

Both honeycomb materials were purchased from Nida-Core and were delivered in

61 cm X 122 cm sheets that were then cut into the final dimensions. The honeycomb

panels were constructed of a 13 mm thick polypropylene core, designated H11-60 PP,

with a thermo fused non-woven polyester cloth face sheet to which either the lauan or

fiberglass face sheets were adhered [30]. Figure 4.1 shows the honeycomb with cloth

face sheet and a cross section of the lauan honeycomb panel.

(a) (b)

Figure 4.1 Honeycomb with Cloth Face Sheet (a) and Lauan Composite Panel (b)

The lauan face sheet, also known as meranti plywood, was 2.7 mm thick and the

fiberglass face sheet, designated as18 oz. wet laminated glass W/R, was 0.74 mm thick.

Nida-Core supplied the material properties of the face sheet materials which are shown in

Table 4.1. For these particular panels, one of the most important material properties is


51/256

28

the strength of the bond between the reinforcing face sheet and the cloth face sheet.

Unfortunately, these values were not supplied by the vendor.

Table 4.1 Honeycomb Face Sheet Material Properties

Property Lauan Fiberglass

Thickness (mm) 2.7 0.74

Tensile Modulus (MPa) 11032 13790

Compressive Modulus (MPa) 11032 15513

Flexural Modulus (MPa) 11032 13445

Tensile Strength (MPa) 6.89 204.8

Compressive Strength (MPa) 6.89 181.3

Flexural Strength (MPa) 6.89 289.6

Shear Strength (MPa) 7.58 96.5

Although the bending stiffness will be dominated by the face sheet properties, the

failure of the panel could be one of several failure modes including failure of the core.

The core properties needed for finite element modeling and failure prediction were only

partially supplied by Nida-Core on their website [31]. The other properties were derived

mainly using analytical formulas found in a widely cited book by Gibson and Ashby

entitled Cellular Solids[32]. Other sources were used as needed to fill in missing

properties and are referenced when presented. In their book, Gibson and Ashby provided

derivations for the material properties of the honeycomb based on the properties of the

solid material, in this case polypropylene. For the analysis included in this research,

three material properties are need for solid polypropylene. These properties are the

density, elastic modulus, and the Poissons ratio. Gibson and Ashby supplied the density

and modulus and the Poissons Ratio waspublished in an additional source. These

properties are shown inTable 4.2.


52/256


53/256

30

necessary, the modulus would be updated to reflect the higher value if a poor match

between the test data and FEM was observed.

In their book, Gibson and Ashby derive the properties of the honeycomb core

using equilibrium equations that relate the properties of the core to the geometry of the

cell and the material properties of the raw material. For the remaining properties needed

for the finite element modeling it was assumed that the cells of the honeycomb core are

regular hexagons which are characterized by equal length sides and interior angles of

120. The first derivations made by Gibson and Ashby relate the in-plane moduli to the

relative densities of the polypropylene and the honeycomb and the polypropylene

modulus as

. (4.2)

Using the provided data and material properties results in an in-plane modulus of

0.6 MPa, which is relatively small modulus especially compared to the face sheets and

even the out-of-plane modulus. Other published research on the mechanical properties of

honeycomb suggests using an in-plane modulus of 0.0 or possibly a very small number in

order to avoid numerical instability [35]. The calculated value of 0.6 MPa is sufficiently

small and yet large enough to avoid numerical issues. This value will be used for both

the in-plane moduli.

There are a total of 6 Poissons ratios for an orthotropic material with only 3

independent. Those needed for the analysis are , , . Gibson and Ashby statethat is equal to the Poissons ratio of the polypropylene, which fromTable 4.2 is 0.42.Utilizing their derivations with the assumption of a regular hexagon results in a value of

1.0 for and applying the reciprocal relation between elastic modulus and Poissons


54/256


55/256

32

4.1.2 Cementitious Plates

A previously developed mix design containing polyvinylbutyral and styrene-

butadiene-rubber (SBR) acrylic latex was selected for the cementitious panels, one with

and one without the addition of 0.6% by volume polyvinyl alcohol (PVA) fibers [36].

This mix was chosen for its relative low modulus and high strength. Several 61 cm x 61

cm plywood molds (Figure 4.2)were built, lined with plastic, and equipped with covers

in order to facilitate making the cementitious panels. The plastic liner was purchased at a

local hardware store and had an adhesive back. The liner was adhered to the bottom

surface, side rails, and cover with special care taken to prevent as many air bubbles as

possible.

Each matrix was mixed using a Hobart mixer and then hand placed in the

honeycomb core. A total of 6 molds were filled; 3 with each matrix. After placing, the

panels were vibrated to reduce the voids that occurred during initial placement. After the

mold was filled, it was covered with a plastic lined lid to prevent shrinkage and the mold

was placed in a humidity chamber so that it could cure for 14 days. The molds were

removed from the chamber and the panels were extracted from the molds. The extraction

from the molds proved to be a challenge.


56/256

33

(a) (b)

Figure 4.2 Honeycomb Core Placed in Mold (a) and Mix Being Hand Placed (b)

The adhesive back on the plastic prevented the liner from first being removed

from the mold and the cementitious panel was stuck to the liner. The bond between the

panel and the plastic liner had to be broken carefully which required the use of several

different tools and several hours. A better solution would have been to adhere the plastic

liner in such a way that it would be easily removed from the molds and then the plastic

could be peeled away from the plates. A chemical mold release was not used to ensure

that no reaction between the release and the polyurea could occur. After removal the

plates were allowed to dry for an additional 48 hours.

After the cementitious composite core dried, both sides were sanded so that a

graphite weave could be bonded to each one of them using epoxy. During sanding, it was

observed that the honeycomb was below the bottom surface of the concrete indicating

that the honeycomb floated in the concrete during placing resulting in a plate thicker

than 13 mm thick honeycomb. After sanding and wiping, a West Systems epoxy was

prepared according the manufacturers directions. The epoxy applied to the surface of

the concrete and to the graphite sheet, then the graphite was placed and the excess epoxy

was removed. This process was repeated for one surface of each panel separately to avoid


57/256

34

making large batches of epoxy, which can lead to premature curing due to the heat

produced during the mixing of the resin and the catalyst. The panels were allowed to

cure for 24 hours and then repeated for the other surface of each panel. The panels were

then allowed to cure for an additional 24 hours before cutting them to produce the final

specimens. A completed panel before cutting is shown inFigure 4.3.

Figure 4.3 Example of Completed Panel Before Final Cutting

A graphite leno weave purchased from Cytec Fiberite, Inc. was used for the

reinforcement. A schematic is shown inFigure 4.4. It consisted of non-impregnated

graphite fibers with 3,000 fibers per tow, spaced at 3.18 mm intervals. Each tow was

0.19 mm thick and 1.07 mm wide. The layer thickness was measured to be 0.381 mm

and the percent of open area was determined to be 44%. According to the manufacturer,

the elastic modulus and tensile strength of the fiber is 231 GPa and 3.65 GPa,

respectively.


58/256

35

(a) (b)

Figure 4.4 Schematic of Graphite Mesh (a) and Detail View with Cell Dimensions (b)

For each of the different panels, three of the four specimens were coated with

polyurea and one was left uncoated. The specimens that were sprayed were coated 1) on

the bottom surface with polyurea, 2) the top surface with polyurea, and 3) coated on both

surfaces. The black aromatic polyurea was exclusively used for all specimens.

4.2 Testing Methodology

In order to conduct the four point bending tests according to ASTM standards, a

test frame was designed and built which is shown inFigure 4.5. The test frame was used

with a 98 kN capacity MTS testing machine. The inner (upper and fixed) and outer

(lower and movable) supports were situated at distances of 15.2 cm and 45.7 cm apart,

respectively; placing the central section in pure bending. A load cell was used to measure

the total force, P that was distributed to both supports and the central span moment could

be easily calculated given the distances between rollers. All tests were run at a loading

rate of 2.54 mm/min.


59/256

36

Figure 4.5 Four Point Bending Setup with Test Frame (Blue)

In addition to load, the MTS machine was capable of measuring the cross head

displacement and this was recorded for each specimen. Referring again toFigure 4.5,the

upper crosshead position was fixed and the lower crosshead was controlled such that the

desired loading rate (2.54 mm/min) was obtained. When recording the crosshead

displacement, it is clear from the figure that this is equivalent to the displacement at the

outer rollers. In a specimen with uniform geometry and therefore a constant moment of

inertia, the displacement at any point along the span can be calculated using well known

flexure formulas with the knowledge of the displacement of one point. Although

attempts were made to produce uniform specimens, the thickness of the cementitious

plates in particular varied across the span leading to a moment of inertia as a function of

span. This fact lead to the use of a gage dial to measure the center deflection relative to

the inner supports as can be seen inFigure 4.5. Unfortunately the dial gage, and


60/256

37

corresponding force, was not recorded by the data acquisition system and had to be read

manually, leading to errors in this measurement.

4.3 Results and Discussion

Moment versus crosshead displacement plots for each material and configuration

are shown inFigure 4.6. Additionally, moment versus center displacement plots are

shown inFigure 4.7. A table summarizing the test results for all 16 specimens is

presented inTable 4.4.


61/256

38

(a) (b)

(c) (d)

Figure 4.6 Moment Versus Crosshead Displacement for (a) No Fiber Cementitious, (b) Fiber

Cementitious, (c) Fiberglass Honeycomb, and (d) Lauan Honeycomb


62/256

39

(a) (b)

(c) (d)

Figure 4.7 Moment Versus Center Displacement for (a) No Fiber Cementitious, (b) Fiber

Cementitious, (c) Fiberglass Honeycomb, and (d) Lauan Honeycomb


63/256


64/256

41

visual inspection of the photographs taken during the tests revealed that, in many cases,

the deflection of the beam did not follow the profile predicted by the elastic curve.

Upon physical inspection of the specimens, the likely reason for the unexplained

results was discovered. The polyurea manufacturer suggested that the specimens first be

cut from the panels and then coated, as opposed to coating the panels first and

subsequently extracting the specimens. The manufacturer rationalized that cutting

through the polyurea coating may change its properties and/or weaken the bond between

the polyurea and the substrate. While spraying the coating on the individual specimens,

however, the edged were left unmasked. The overspray varied significantly along the

length of the span and from specimen to specimen. As explained in the paragraphs that

follow, this selectively reinforced the bond between the core and the substrate which led

to many unexpected developments.

Table 4.5 includes a set of notes regarding the deposition of polyurea on the edges

of the specimens and a description of the failure mode that occurred in each.Figure 4.8

includes photographs to clarify the nomenclature used to describe polyurea deposition.

Table 4.5 Notes Regarding Polyurea Deposition and Failure Mode.

No.Specimen Description and Polyurea

Deposition Notes Regarding Failure Mode

1 No Fiber Uncoated: no overspray.

Panel failed in tension at bottom within

center span.

2

No Fiber Black on Bottom: significant

overspray on bottom edges, front and

back; dots of polyurea on top edges,

front and back.

Panel delaminated on top within center

span. Separation occurred betweengraphite/epoxy layer and concrete. Some

regions along delamination remained

connected by polyurea dots. In these areas,

concrete failed in compression.

3

No Fiber Black on Top: significant

overspray on top edges, front and

back; regions of polyurea on bottom

Panel buckled on top within center span.

Polyurea bridged gap underneath

graphite/epoxy layer. A local buckle formed


65/256

42

edges, front and back. prior to failure.

4

No Fiber Black on Both; overspray very

heavy on front side; significant

overspray on top edge, back side;regions of polyurea on bottom edge,

back side.

Panel buckled on top within center span.

Polyurea bridged gap underneathgraphite/epoxy layer. A local buckle formed

prior to failure.

5 Fiber Uncoated: no overspray.


span.

6

Fiber Black on Bottom: significant

overspray on bottom edges, front and

back; dots of polyurea on top edges,

front and back.


span.

7

Fiber Black on Top: overspray very

heavy on front side; significant

overspray on top edge, back side; nopolyurea on lower edge, back side.

Panel failed at bottom in tension withincenter span.

8

Fiber Black on Both: overspray very

heavy on front side; no significant

overspray on edges, back side.


span on the side that had no polyurea.

9

Nida Fiberglass Uncoated: no

overspray.

Panel delaminated on top outside of center

span.

10

Nida Fiberglass Black on Bottom:

regions of polyurea on bottom edges,

front and back; regions of polyurea on

top edge, back side; no polyurea on top

edge, back side.

Panel delaminated on bottom outside of

center section along the edge where there

was no polyurea.

11

Nida Fiberglass Black on Top: regions of

polyurea on top edge, front side;

significant overspray on top edge, back

side; regions of polyurea on bottom

edges, front and back


center span.

12

Nida Fiberglass Black on Both:

overspray very heavy on front side;

regions of polyurea along top and

bottom edges, back side.


center span.

13 Nida Lauan Uncoated: no overspray.


center span.

14

Nida Lauan Black on Bottom: dots of

polyurea on bottom edges, front and

back; no polyurea on top edges, front

and back.

Panel delaminated on top outside of center

span.

15

Nida Lauan Black on Top; overspray

very heavy on top edge, back side;

regions of polyurea on top edge, front


center span.


66/256

43

side; no polyurea on bottom edges,

front and back.

16

Nida Lauan Black on Both: overspray

very heavy on both sides.

Panel delaminated on bottom within center

span.

(a) (b)

(c)

Figure 4.8 Photos and Examples of Nomenclature Used to Describe Polyurea Deposition: Heavily

Coated on Front Side (a); Significant Overspray on Top Edge, Regions of Polyurea on Bottom Edge,

Front Side (b); Dots of Polyurea on Both Edges, Front Side (c)

All but two of the 16 bending specimens failed with some type of delamination of

the reinforcement (cementitious panels) or face sheets (honeycomb panels). Only

Specimen 1 and Specimen 7 failed in a different manner when they failed abruptly in


67/256

44

tension as the reinforcement fractured. All of the other specimens failed when the

adhesive between the reinforcement or face sheets failed in shear or when the

reinforcement or face sheets buckled due to the compressive loads. It was observed that

when compared to its uncoated counterpart, specimens with significant overspray along

structural enhancement using polyurea

Documents